CORAL REEFS, AND ISLANDS. 279 



relegated to the category of those many hypotheses 

 which have, indeed, helped science for a time, by 

 promoting and provoking further investigation, but 

 which, in themselves, have now finally ' kicked the 

 beam.' 



" But this great lesson will be poorly learnt unless 

 we read and study it in detail. What was the flaw in 

 Darwin's reasoning, apparently so close and cogent ? 

 Was it in the facts, or was it in the influences ? His 

 facts in. the main were right ; only it has been found 

 that they fitted into another explanation better than 

 into his. It was true that the corals could only grow 

 in a shallow sea, not deeper than from twenty to 

 thirty fathoms. It was true that they needed some 

 foundation provided for them, at the required depth. 

 It was true that this foundation must be in the pure 

 and open sea, with its limpid water, its free currents, 

 and its dashing waves. It was true that they could 

 not flourish, or live in lagoons, or in channels, however 

 wide, if they were secluded and protected from 

 oceanic waves. One error, apparently a small one, 

 crept into Darwin's array of facts. The basis, or 

 foundation, on which corals can grow, if it satisfied 

 other conditions, need not be solid rock. It might 

 be deep-sea deposits, if these were raised or elevated 

 near enough the surface. Darwin did not know this, 

 for it is one of his assumptions that coral ' cannot 

 adhere to a loose bottom.' The Challenger obser- 



